Oxygen Toxicity

GaAs epitaxial layers have been grown by close-spaced vapor transport (CSVT) from various p-type Zn doped GaAs sources, using H20 as transporting agent. The doping impurity concentrations in the sources were in the 10 TM, 10 I8, 1019, or 1020 cm -~ ranges. Heavily compensated n-type epitaxial layers were obtained with sources characterized by an acceptor 16 18 concentration in the i0 -I0 range, while p-type layers were obtained with sources having Zn concentrations equal or greater than i0 TM cm -3. About 1% of the Zn present in the source is found to be electrically active in the epitaxial layer. With the help of a diffusion control model based only on thermodynamic analysis, it is possible to demonstrate that Zn is transported as ZnO. Using the same model, it is also shown that Te, a n-type doping impurity characterized in CSVT by a transfer coefficient of unity, is transported as Te2 under the same growth conditions.Close-spaced vapor transport (CSVT) is a special kind of chemical vapor deposition (CVD) technique developed about 25 years ago almost simultaneously by Sirtl and a RCA team (1-5). Its particularity lies in the displacement of the thermal equilibrium of the transport reaction resulting from different source and substrate temperatures. This provides the driving force of the transport. During the deposition, the source and substrate are close together in the reactor. The transporting agent, such as H20 used in the transport of GaAs, is introduced into the reactor with the help of a carrier gas such as H2. All the problems relative to the storage and use of highly toxic reagents necessary in conventional CVD of GaAs are therefore avoided in CSVT.It has been demonstrated that GaAs wafers can be used as sources in CSVT to grow epitaxial layers of GaAs on GaAs substrates (6-11) and also on Ge substrates (1-4). However, in order to produce by CSVT solid-state_ devices like solar cells, it is necessary to control the electrical properties of the grown layers. An attempt toward this goal is the use of doped GaAs wafers as the source of reagents. In this case, As2 (or As4) and Ga20 obtained by the reaction of GaAs with H20 (the transporting agent) are provided along with the doping impurity to be incorporated in the growing layer. It has been shown (12) that n-type Te-doped GaAs sources can be used to grow n-type doped layers with a Te transfer coefficient equal to unity. By comparison (11), layers grown from undoped semi-insulating GaAs sources are also n-type but with a background doping density in the low 1016 cm -3. The behavior of Si-doped GaAs sources has also been investigated (12). In that case, the oxidation of Si by water impedes its transfer. SiOx accumulates at the surface of the source where it can be detected by Auger spectroscopy.It is the purpose of this work to demonstrate that p-type GaAs layers can be obtained by CSVT using Zn doped GaAs sources and H20 as transport agent. However, it is shown that only -1% of the Zn present in the source is found to be electrically active in the epitaxial...

Section: Discussionmentioning

confidence: 99%

“…The same assumption can be made for the equations involving the transport of the doping impurities. For instance, as far as reaction [4] is concerned, one has Jz~ = Jz,o-Dz,o.az [rK(TI____~)] nz___% [12]…”

Section: ( Jmentioning

confidence: 99%

Electrical Characterization of GaAs Epitaxial Layers Grown by CSVT from Zn‐doped GaAs Sources

Cossement

Dodelet

Bretagnon

et al. 1991

“…The close-spaced technique can also be used with reversible decompositions. The reaction usually proposed for the transport of GaAs is (22,23,31) 2 GaAs(s~ + HzO(g> ~ Ga20(g) + Asz(g) + H2(g> [3] For the epitaxial growth of GaAs on GaAs, this reaction is happening at two places in the system: (i) on the sur- The model developed to calculate the growth rate is based on the following hypothesis. (i) During the growth of epitaxial GaAs, the concentrations of volatile products (Gala, As~) are always constant at the source surface at T1, and at the substrata surface at T~.…”

Section: Discussionmentioning

confidence: 99%

“…GaAs, GaP, and GaAsxPl_x films have also been grown by CSVT in the early days of this technique (22)(23)(24)(25)(26)(27)(28) and then again quite recently (29)(30)(31). These studies have shown the important role of water vapor usually used as transporting reagent, although I2 or ZnC12 have also been proposed for the same purpose.…”

mentioning

confidence: 99%

Epitaxy of GaAs by the Close‐Spaced Vapor Transport Technique

Côté

Dodelet

Lombos

et al. 1986

Thin epitaxial normalGaAs films have been grown on (100) semi‐insulating normalGaAs (SI normalGaAs ) by close‐spaced vapor transport (CSVT). The epitaxy has been confirmed by x‐ray diffraction spectra and SEM observation of etch pit shapes and orientations. The effect of hydrogen flux, water vapor pressure, distance between the source and the substrate, and transport temperature on the film growth has been measured and analyzed. A model based on the variation of the equilibrium constants of the reactions occurring at the surface of the source and the substrate has been developed to elucidate the growth of normalGaAs . The measurements of transport properties, performed on the CSVT depositions, show that the deposited normalGaAs films are always n‐type independently of the used source (n‐type, p‐type, or SI normalGaAs ). The unintentionally introduced dopant concentration may vary from 5×1016 normalto 2×1018 cm−3 . It is shown that shallow traps associated to EL2 deep levels can be responsible for part of this doping, on the lowest charge carrier density side, but not above ∼5×1017 cm−3 . An unwanted dopant is always introduced by the present technique. It is proposed that Si could be the unwanted doping agent.

“…Saraie et al have reported the epitaxial growth of CdTe by a close-spaced technique (7). More recently, J. Mimila-Arroyo et al (8,9) made a detailed study using water vapor as the transporting agent in the same way as R. F. Lever and F. Jona (10) for germanium growth and G. Gottlieb and J. Carboy (11) for GaAs deposition [cf. (2,3)].…”

Section: Application To Cdte Transportmentioning

confidence: 99%

Simplified Theory of Reactive Close‐Spaced Vapor Transport

Bailly

Cohen‐Solal

Mimila-Arroyo

1979

A simplified model is proposed to determine the thermodynamic constants of the transport reaction in a close-spaced configuration, even when the reacting species are unknown. For example, the enthalpy variation, the activation energies of condensation, and sublimation in the water vapor transport of CdTe are determined (AHeq. = 22, Qcond. ----9.2, and QsubL = 31.2 in kcal/ mole-D. The model is useful for interpreting the experimental results in epitaxial processes. In the low pressure iodine transport of silicon, it is shown that the growth rate is not a diffusion controlled process; from the energy AHeq : --16.5 kcal/mole -1 for the reaction Si + 2I ~--SiI2, we found Qcond. "----8 and Qsubl. : --24.5 (kcal/mole-D. An important result of the analysis is that determination of enthalpy variation cannot be obtained only from the slope of the growth rate-temperature dependence.The use of a close spacing in chemical transport systems for growing epitaxial layers of pure or compound semiconductors has been emphasized by Sirtl (1), Robinson (2), and Nicoll (3). The most important features of the technique are the use of a close spacing between source and substrate (1 mm) between which is maintained a temperature gradient, and the fact that a gas is used to react with the source to form a volatile compound which is decomposed at the surface of the substrate. Generally, the transport reaction occurring in the closed spaced system is difficult to establish, and it is the aim of this paper to propose a simple theoretical model for the mass transfer even if the transporting species are unknown. Although the deposition rate is strongly sensitive to the surface preparation and the crystallographic orientation of the substrate (4, 5) as well as to the pressure, or the flow rate, of the reactive gas, the substrate and source temperatures are of major importance in the process only in so far as the single crystallinity of the layer is not concerned. The proposed model is shown to be useful for endothermic as well as exothermic process, i.e., for positive or negative heat of reaction at constant pressure.